β-Blockers:  The New Standard of Therapy for Mild Heart Failure FREE

Over the past decade, large-scale, randomized clinical trials have shown that β-blockers improve cardiac performance, slow disease progression, reduce hospitalization for cardiovascular and all causes, and—in the cases of bisoprolol, long-acting metoprolol, and carvedilol—improve survival in patients with chronic systolic heart failure.^1- 2 Despite these proven therapeutic benefits, many physicians are reluctant to prescribe β-blockers to patients with mild heart failure, especially when first-line medications (diuretics, digitalis glycosides, and angiotensin-converting enzyme [ACE] inhibitors) seem to be effective at alleviating symptoms. Several questions or concerns may underlie this hesitation, the foremost being "Why prescribe additional drugs when a patient's symptoms seem to be under control?" There also may be some lingering concern about the former view that β-blockers were contraindicated for the treatment of heart failure. Finally, the busy clinician may wonder if the potential benefits of β-blockade warrant the extra care required during initiation of treatment, particularly among patients who are responding to conventional therapy.

The following review will attempt to answer these questions by discussing the underlying pathophysiological characteristics of heart failure, the clinical benefits of administering β-blockers during the early stages of systolic heart failure, and the practical considerations of initiating therapy. The evidence presented hereafter suggests that the optimal treatment for systolic heart failure, especially in patients with mild symptoms, includes a β-blocker in addition to conventional medications (diuretics and digoxin) and an ACE inhibitor.

The sympathetic nervous system becomes activated early in the course of left ventricular (LV) systolic dysfunction regardless of its cause, and remains activated throughout the natural history of clinically overt heart failure.³ Initially, adrenergic activity compensates for the failing heart. That is, norepinephrine stimulates ventricular contraction, modifies vascular resistance, and redirects blood flow to central organs, and thus bolsters cardiac output and blood pressure.^4- 5 Over the long term, however, heightened sympathetic tone injures the heart in a progressive fashion. In fact, sustained adrenergic activation in the myocardium, kidneys, and peripheral vasculature seems to be one of the primary mechanisms for the worsening of heart failure or disease progression, as indicated by the well-established correlation between high levels of plasma norepinephrine and poor long-term prognosis.^6- 7

The adverse effects of chronic sympathetic activation are numerous, complex, and interdependent. Elevation of cardiac output, for example, taxes the heart and increases the demand for oxygen in the myocardium, setting the stage for both ischemia and oxidative stress. Meanwhile, peripheral arterial and venous vasoconstriction increases ventricular afterload and preload, placing an additional burden on the failing pump.⁸ Over time, sustained mechanical stress and the direct biological effects of norepinephrine (eg, myocyte hypertrophy and programmed cell death^9- 13) cause cardiac remodeling, which is characterized by morbid changes in ventricular geometry, mass, and volume, and which results in a dilated and less contractile chamber.¹⁴ Ischemic injury in the oxygen-hungry, dilated ventricle kills tissue and further reduces contractility, and in turn leads to additional dilation.¹⁴ Ultimately, successive rounds of ischemia, tissue death, and compensatory dilation may result in myocardial infarction or complete ventricular failure.¹⁵

Other adverse consequences of chronic sympathetic activation include tachycardia and desensitization of β-receptors in the myocardium, both of which undermine LV function by further decreasing contractile strength.^16- 17 In addition, sustained stimulation of adrenergic receptors unfavorably affects the transport and/or homeostasis of various ions (notably sodium, potassium, and calcium),^18- 19 and thereby disturbs the timing and coordination of myocardial muscle contraction. The resulting arrhythmias may play an important role in triggering sudden cardiac death in patients with heart failure.²⁰

As shown in Figure 1, chronic adrenergic activation is part of a vicious cycle that leads to progressive LV dysfunction. It is important to note that this pathophysiological process begins before the symptoms of heart failure emerge.²¹ By the time the syndrome reaches its mildly symptomatic state, progressive damage is well under way. This observation underscores the need to diagnose and optimally treat heart failure in its earliest stages.

Systolic or diastolic LV dysfunction may underlie the clinical syndrome of heart failure; ie, progressive LV damage and clinical symptoms may ensue from diminished myocardial contractility or from abnormalities of diastolic ventricular filling, respectively. Most patients (≈70%) with heart failure have primarily systolic dysfunction, sometimes accompanied by diastolic dysfunction. Coronary artery disease is the cause of heart failure in about two thirds of patients with systolic dysfunction; patients may also have nonischemic causes of cardiomyopathy such as hypertension, valvular heart disease, myocarditis, systemic disease, toxins, alcohol/drug use, or idiopathic cardiomyopathy.²² Of note, the Studies of Left Ventricular Dysfunction (SOLVD) registry includes only patients with systolic heart failure of predominantly ischemic cause (70% patients),²³ while the Framingham Study,²⁴ which observed patients for more than 40 years, is made up of patients with either systolic or diastolic heart failure; hypertension with or without coronary artery disease is the cause of heart failure in nearly 70% of patients in the Framingham Study.

The hallmark of systolic dysfunction (and what also differentiates it from diastolic failure) is a depressed LV ejection fraction (≤40%), which may also be the ideal marker of disease progression. For systolic heart failure, the goals of treatment are to slow progression of disease, conferring a reduced risk of morbidity and death, and to improve symptoms by treatment of volume overload. The balance of alleviating symptoms while managing disease progression through treatment approaches that block neurohormonal stimulation is the current aim of the medical regimen.²²

Diastolic dysfunction may be caused by coronary artery disease, hypertension, diabetes mellitus, aortic stenosis, hypertrophic cardiomyopathy, infiltrative cardiomyopathies, and endocardial fibroelastosis. The prevalence of diastolic heart failure correlates with age; decreasing ventricular compliance with increasing age may be responsible for diastolic disease later in life.²⁵ Diastolic dysfunction is defined as normal or preserved rest systolic function in the presence of heart failure signs and/or symptoms.²⁶ Because there are no well-controlled, randomized, large-scale trials in patients with diastolic heart failure, treatment strategies are based on empirical data. Symptoms of congestion should be managed, ischemia prevented, and, if possible, the underlying causes identified and treated.²² The remainder of this review focuses on the management of systolic heart failure with particular attention to the role of β-blockers.

Current first-line treatments for heart failure either overlook or incompletely inhibit chronic adrenergic activation. Diuretic-based therapy, the most widely used form of treatment, corrects the hemodynamic disturbances that accompany heart failure (eg, edema and congestion), but does not reduce (and may, in fact, increase) the activity of the major neurohormonal vasoconstrictor systems that fuel the progression of the syndrome.²⁷ As such, the improvements associated with diuretic treatment are transitory and ultimately misleading; ie, they are improvements in symptoms alone. Although digitalis glycosides modestly reduce cardiac sympathetic tone and improve heart function, they have a limited effect on the natural history of the disease. For example, a recent placebo-controlled study involving more than 7000 patients with chronic heart failure showed conclusively that digoxin does not improve survival.²⁸ Angiotensin-converting enzyme inhibitors counter the ill effects of the renin-angiotensin system (one of the other major neurohormonal systems involved in the pathophysiological characteristics of heart failure), and thereby both improve symptoms and significantly reduce mortality.²⁹ Unfortunately, rates of hospitalization and death remain high among patients with heart failure treated with an ACE inhibitor in addition to diuretics and digoxin,³⁰ indicating that even the most robust form of standard therapy does not offer optimal protection against progressive cardiac dysfunction. The following section reviews clinical evidence supporting the notion that inhibiting chronic adrenergic activation via β-blockade attenuates disease progression and reduces morbidity and mortality in patients with heart failure.

The primary concern guiding the decision to add a new agent to an existing regimen is practical in nature: will the new agent augment clinical benefits? Most importantly, will the added therapy improve the patient's health and reduce the risk of hospitalization and death? Over the past decade, the results of numerous randomized, controlled clinical trials have demonstrated that β-blockers both improve the symptoms of systolic heart failure and, most importantly, impede disease progression when added to conventional therapy. Long-term (≥3 months) use of β-blockers in combination with diuretics and an ACE inhibitor (and sometimes digoxin) consistently and significantly increases LV function, as measured by ejection fraction,^31- 35 and reduces the incidence of hospitalization in patients with a broad range of clinical symptoms^{1- 2,31,36- 37} (Table 1).

The recently published Consensus Recommendations for the Management of Chronic Heart Failure,³⁸ as well as 3 meta-analyses of over 3000 patients in more than 20 randomized, double-blind, placebo-controlled trials^39- 41 demonstrating a beneficial effect of β-blocker treatment in chronic heart failure, support and provide guidelines for early β-blocker initiation and long-term use in heart failure as part of current therapeutic management. In fact, because the development of symptoms is only weakly associated with the evolution of cardiac dysfunction, the goal should be to minimize and prevent disease progression by employing β-blocker therapy early in the antihypertensive/heart failure regimen with diuretics and/or ACE-I treatment.^38,42

With the extensive data on carvedilol, the only β-blocker approved by the Food and Drug Administration for the management of systolic heart failure, the results of The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II)² and preliminary results from the Metoprolol CR/XL [controlled release/extended release] Randomized Intervention Trial in Congestive Heart Failure (MERIT-HF),⁴³ there are sufficient data on patients with mild to moderate heart failure (New York Heart Association [NYHA] functional classification II-III) to provide strong evidence that β-blockers provide additional benefit in these patients. Therefore, physicians should not delay adding treatment with β-blockers to patients' current heart failure regimens.

Evidence for the clinical advantage of initiating β-blockade during the early (mild) stages of systolic heart failure is derived, in part, from the findings of the US Carvedilol Heart Failure Trials Program (US-CHFTP).³⁶ In these studies, mild heart failure is described as the ability to walk between 425 or 450 and 550 m during a 6-minute corridor walk test; the NYHA class is determined; and the left ventricular ejection fraction (LVEF) is measured by radionuclide ventriculography. In a placebo-controlled component study of the program, Colucci and colleagues³⁶ administered carvedilol—a nonselective β-blocker with vasodilatory, antioxidant,^44- 45 and antiproliferative^46- 47 properties—to patients with mild symptoms of heart failure. During an average treatment period of 6.5 months, addition of carvedilol to the standard regimen (diuretics and an ACE inhibitor, with or without digoxin) resulted in a 48% reduction in clinical progression (P=.008 vs placebo), defined as death due to heart failure, hospitalization for heart failure, or the need for a sustained increase in heart failure medications (Figure 2). Several secondary end points showed improvement as well. As in earlier studies,^48- 50 carvedilol significantly increased the LVEF, a measure of systolic heart function with strong prognostic value.⁵¹ Carvedilol also improved various indices of clinical status, including NYHA functional classification and heart failure symptom score. Importantly, for each of these clinical indices, carvedilol reduced the percentage of patients whose condition worsened over the course of the trial (Table 2), further indicating that the addition of carvedilol to the standard regimen impeded disease progression. Similar to bisoprolol-treated patients in CIBIS-II, carvedilol significantly increased survival over placebo in patients receiving treatment; as seen by the divergence of the Kaplan-Meier curves (Figure 3), the beneficial effect was evident early in therapy and increased gradually to the end of the trial.^1- 2 Overall, the results of this study underscore the importance of initiating β-blockade to slow disease progression even when the symptoms of heart failure do not seem to warrant additional therapy. These observations are further supported by the results of the Australia-New Zealand carvedilol study of mild heart failure.⁵²

The CIBIS-I investigators³⁵ evaluated bisoprolol, a highly selective β₁-adrenergic receptor blocker; β₁-receptors are found mainly in the ventricle of the heart. The results of CIBIS-I demonstrated a nonsignificant trend toward lower mortality (20%) in the bisoprolol group and 30% fewer hospital admissions for worsening heart failure (P<.01). Based on these results, CIBIS-II was designed to further evaluate this evidence; CIBIS-II data indicate that bisoprolol both reduces morbidity and significantly improves survival when added to standard therapy.² Among 2647 patients with moderate to severe symptoms of heart failure, bisoprolol produced a 34% reduction in overall mortality (P<.001), a 44% reduction in sudden deaths (P=.001), a 36% reduction in hospitalization for heart failure (P<.001), and a 20% reduction in all-cause hospitalization (P<.001) compared with placebo during an average treatment period of 1.4 years.

Metoprolol is also a cardioselective β-blocker that has been shown to improve left ventricular function and symptoms of heart failure, and to decrease the number of hospitalizations due to heart failure. The MERIT-HF study⁵³ was designed to investigate the effect of once-daily dosing of controlled release/extended release metoprolol treatment when added to standard therapy in patients with mild to severe disease. Primary objectives include all-cause mortality and all-cause hospitalizations (time to first event). The secondary study end point was the combined risk of total mortality and hospitalizations for worsening heart failure; tertiary end points assessed NYHA functional class, tolerability, and health economics. After a 2-week placebo run-in period, 3991 clinically stable patients (LVEF ≤0.35) were randomized to placebo or long-acting metoprolol and titrated slowly to a target dose of 200 mg daily. The MERIT-HF study was terminated early by the independent safety committee after reaching the preplanned end point of reduced risk of total mortality (35%; P<.001).⁴² The CIBIS-II and MERIT-HF data are important because they demonstrate a significant and unequivocal effect of β-blockade on survival per se (rather than in combination with other clinical end points such as hospitalization), and because they reinforce earlier data suggesting that combination therapy including a β-blocker affords the best available protection against progressive systolic dysfunction.

The Beta-blocker Evaluation of Survival Trial protocol of moderate to severe heart failure was recently completed, and preliminary results indicate an approximate 10% (nonsignificant) reduction in all-cause mortality with bucindolol (presentation at the American Heart Association annual scientific sessions, Atlanta, Ga, November, 1999). However, in patients who were similar to those studied in US-CHFTP, CICIS-II, and MERIT-HF (predominantly white and class III), bucindolol significantly reduced mortality by 22%.⁵⁴ The Carvedilol Prospective Randomized Cumulative Survival study is an ongoing placebo-controlled trial with all-cause mortality as the primary end point in 2500 patients with moderate to severe heart failure (NYHA IIIb-IV) observed for at least 16 months; results are expected to be available in 2002. These 2 studies when completed or completely reported should definitively describe the mortality risk in patients with both ischemic and idiopathic cardiomyopathy and moderate to severe disease treated with nonselective β-blockers.

With the exception of the Metoprolol in Dilated Cardiomyopathy study,³³ which evaluated nonischemic patients with mild to moderate systolic heart failure, the recent large-scale studies evaluating β-blocker therapy in systolic heart failure assessed patients with both ischemic and nonischemic cardiomyopathy. The US-CHFTP and CIBIS-II as well as meta-analyses of β-blocker trials found that treatment effects were independent of the severity or cause of disease.^41,55

Only the US-CHFTP and MERIT-HF included patients with mild disease (NYHA II). In fact, patients enrolled in most of the major β-blocker trials generally had class III symptoms, and nearly all of the patients were receiving stable doses of background therapy with digitalis, diuretics, and an ACE inhibitor. Furthermore, most of the large-scale β-blocker studies did not enroll patients without any symptoms (class I), and only 3% to 5% of patients enrolled in the Metoprolol in Dilated Cardiomyopathy study, CIBIS-I, and US-CHFTP had class IV symptoms. The Australia-New Zealand Trial⁵² was the only protocol in which almost one third of the patients were class I at randomization; nearly two thirds were not receiving digitalis.^39- 41

Of the 4 component protocols that make up the US-CHFTP, 97% of patients were characterized as having class II or III symptoms and were receiving conventional therapy at study entry. Overall mortality in the US-CHFTP was 1.9% vs 5.9% in mild/class II and 4.2% vs 11% in moderate/class III patients with heart failure randomized to carvedilol or placebo, respectively.^1,56

Meta-analyses of β-blocker randomized trials in which most of the patients had NYHA class II or III symptoms despite the use of triple background therapy found that β-blockade decreased the risk of all-cause mortality by 30% to 32% (P=.003); the annual mortality rate for placebo was approximately 11%. This survival effect was similar for class II and class III patients.^39- 41 These results are consistent with the magnitude of the survival effect of ACE inhibitors that have been evaluated in over 7000 patients in more than 30 placebo-controlled trials (16%-34% [average, 26%] reduction in mortality). All trials enrolled patients with systolic dysfunction (LVEF <35%-45%) who were also being treated with stable doses of digitalis and diuretics. The Cooperative North Scandinavian Enalapril Survival Study⁵⁷ enrolled NYHA class IV patients with coronary artery disease (>70%), who experienced a 27% reduction in all-cause mortality (P=.003). In the Vasodilator–Heart Failure Trial II,⁵⁸ comparing enalapril (n=403) with hydralazine and isosorbide dinitrate (n=401), approximately 95% of patients had class II or III heart failure; enalapril decreased mortality by almost 34% at 1 year and 28% at 2 years (P=.02). In the SOLVD Prevention trial,⁵⁹ approximately two thirds of patients in both the enalapril (n=2111) and placebo (n=2117) groups were classified as having class I heart failure and one third had class II; enalapril had no substantial effect on mortality alone, but significantly reduced the risk for first hospitalization for heart failure by 36% (hospitalization for heart failure was associated with an approximate 33% increase in the risk of death). In the SOLVD Treatment trial,³⁰ patients in the enalapril (n=1285) and placebo (n=1284) groups were classified as follows: class I (11%), class II (57%), and class III (30%-31%); less than 2% of patients in each group had class IV heart failure. Patients experienced a 16% reduction in the risk of all-cause mortality.^29,60

In the placebo arms of the Prospective Randomized Amlodipine Survival Evaluation (class III-IV),⁶¹ the Prospective Randomized Milrinone Survival Evaluation (class III-IV),⁶² and the Digitalis Investigation Group (class II-III)²⁸ trial in which all patients in the placebo groups were receiving ACE inhibitors, the more severely symptomatic patients experienced a similar reduction in mortality as the class IV patients of the Cooperative North Scandinavian Enalapril Survival Study trial. In comparison, the mild to moderately symptomatic patients in the ACE/placebo arm of the Digitalis Investigation Group trial experienced a similar survival effect as that observed in the randomized controlled studies of ACE inhibitors.

In several large-scale β-blocker trials, significant reductions in mortality were not convincingly demonstrated because of the combined end points or other limitations of the studies. The following section will review some of these aspects of trial design (Table 1).

Some features of the US-CHFTP remain controversial^63- 64: the run-in period prior to randomization, lack of effect on exercise tolerance, patient selection, and abbreviated duration of treatment. It is a well-accepted design feature of many trials evaluating survival in heart failure³³ to incorporate an open-label, run-in phase prior to randomization to maintain blinding if an active drug is suspected of causing frequent, early adverse effects. The open-label, run-in period also increases the power of the study because the trial is more likely to detect a true survival effect if the patients randomized to active treatment have a greater likelihood of being maintained on the therapy for the entire evaluation time frame. And finally, this design most closely represents clinical practice because only those patients who tolerate the drug are offered the option of long-term treatment.

On the other hand, the drawback of the open-label, run-in design is the interpretation of the analyses should any deaths occur during this period without a parallel control group, ie, whether the risk of death during the run-in phase is the result of the underlying disease or related to the study drug.⁵⁶ The mortality rate during the run-in period of the US-CHFTP was low (0.58%) and similar to that of the SOLVD study open-label phase (0.51%). Furthermore, the mortality rate during the 2-week run-in phase was less than the mortality rate observed in the 3-week screening period preceding the run-in (1.7%), when patients were being assessed for their eligibility to enroll in the program while continuing treatment with their usual heart failure medications. And finally, including the 7 patients who died during the open-label, run-in phase in the analysis still results in a significant reduction in mortality (P=.01). Even if the 7 deaths and 17 episodes of heart failure that occurred prior to randomization were attributed to carvedilol, the results would still stand—carvedilol treatment reduced mortality by 48% (P=.01) and reduced morbidity and mortality by 27% (P=.03).⁶⁶

The measurement of exercise tolerance has been particularly difficult for the evaluation of β-blockers in heart failure because these agents are known to attenuate any favorable effect of therapeutic interventions on exercise performance, especially at peak effort. For this reason, many trials have incorporated methods for assessing submaximal performance as better reflecting patients' daily physical activity; unfortunately, the methods used have not been standardized and are troublesome to interpret.^67- 68

Those enrolled in the trial were by design stable patients with heart failure. Exclusion criteria specified a major cardiovascular event or a surgical procedure within 3 months of study entry or other cardiac symptoms and/or disorders that might put patients at risk for deterioration while initiating and maintaining carvedilol treatment. Patients were required to have chronic heart failure for at least 3 months with stable hemodynamic derangement despite at least 2 months of treatment with diuretics, ACE-I with or without digoxin, hydralazine, and nitrates prior to enrollment.¹ This unchanged therapeutic regimen was required for at least 1 month prior to study enrollment. The CIBIS-II and MERIT-HF programs employed similar eligibility criteria defining patients who were clinically and hemodynamically stable while undergoing conventional therapy at study entry. The added benefit of β-blockade was observed in patients already receiving conventional treatment, particularly ACE inhibitors, suggesting that combined blockade of 2 neurohormonal systems (renin-angiotensin and sympathetic nervous systems) can produce potentially synergistic effects. Importantly, the Consensus Recommendations for the Management of Chronic Heart Failure³⁸ provide guidelines for patient selection specifying that patients with acutely decompensated heart failure should receive treatment for the immediate condition and then be reevaluated for β-blockade after fluid overload has been adequately managed and clinical stability has been achieved.

The CIBIS-II trial, which was stopped early because of the finding that all-cause mortality was significantly lower with bisoprolol treatment, observed patients for a mean of 16 months. The MERIT-HF trial was also prematurely stopped because of a significant survival advantage with metoprolol controlled release/extended release treatment with most patients observed for 24 to 30 months (mean, >12 months). Most of the data on patients receiving carvedilol treatment for more than 7 months are from the Australia-New Zealand protocol involving patients with mild, ischemic cardiomyopathy. Although this trial continued for 15 to 24 months (mean, 19 months), carvedilol decreased mortality by only 28%⁵²; however, the 95% confidence intervals overlapped with that of the US-CHFTP trials. It is of note that the Australia-New Zealand protocol used a lower dose of carvedilol in a formulation with less bioavailability than the drug used in the US-CHFTP protocols.⁶⁶ In the 2 carvedilol component protocols that specifically enrolled 781 patients (366 US-CHFTP patients; 415 patients from Australia-New Zealand) with mild to moderate heart failure, the primary end point was clinical progression of disease. Carvedilol treatment decreased the risk of clinical progression by 48% (P=.008) in the US-CHFTPcomponent and by 26% (P=.02) in the Australia-New Zealand trial. Further, the Australia-New Zealand study demonstrated a 23% decrease in the risk of hospitalization for any reason (P=.05), while the US-CHFTP study resulted in a 77% reduction of all-cause mortality (P=.05).^36,52

Of course, a larger database would strengthen the argument for the positive survival benefit of carvedilol treatment. However, in light of the results of CIBIS-II and MERIT-HF, it is difficult to ignore or minimize the contribution of β-blocker therapy for patients with systolic heart failure and the substantial survival and symptoms benefit demonstrated by the US-CHFTP.

β-Blockers vary widely in their pharmacological properties. As a consequence, one may wonder which agent is most effective. Although it is clear that second- and third-generation agents (eg, metoprolol and bisoprolol, and carvedilol and bucindolol, respectively) have been studied more extensively in heart failure than first-generation drugs (eg, propranolol and timolol), it is not definitively known whether one β-blocker among the recent group is superior to the rest. Some evidence suggests that third-generation agents, in particular carvedilol, which combine nonselective β-blockade with ancillary properties such as vasodilation and protection against oxidation, may produce more comprehensive clinical effects than second-generation drugs, which generally have a single action (selective β₁-blockade). For example, compared with metoprolol, carvedilol exerts a more complete antiadrenergic effect by blocking all 3 adrenergic receptors present in the failing human heart.⁶⁹ In addition, carvedilol, but not metoprolol, improves renal hemodynamics in patients with chronic heart failure.⁷⁰ Other data, such as the CIBIS-II results,² suggest that the absence of potentially advantageous ancillary properties is not a hindrance to successful therapy. Thus, definitive statements about the relative efficacy of one β-blocker vs another await the completion of large-scale, direct comparative studies. In the meantime, one deciding factor may be ease of administration. Carvedilol is unique in this respect in that it is currently the only β-blocker appropriately dosed and formulated for the treatment of heart failure (ie, it is the only agent of its class approved in the United States for this indication).

According to the Consensus Recommendations for the Management of Chronic Heart Failure,³⁸ β-blockers should be added to preexisting treatment with diuretics and an ACE inhibitor and may be used together with digitalis or vasodilators in all patients with stable class II or III heart failure due to LV systolic dysfunction (low ejection fraction) who tolerate the drugs. Treatment should not be delayed until the patient is found to be resistant to other heart failure drugs because these patients are at considerable risk for death, which might be prevented if treatment with a β-blocker is initiated earlier. All ACE inhibitors should be used in a similar fashion and not delayed in patients with mild, moderate, or severe systolic heart failure. However, ACE inhibitors should not be prescribed without diuretics in patients who have a recent history or require management of fluid retention; ACE inhibitors are generally added to treatment with diuretics and may be used together with digitalis or β-blockers.

Most patients remain symptomatic despite treatment; therefore, if digoxin improves clinical status without adversely affecting long-term outcome, it should be used at any time when symptom relief is required. Digoxin is effective across a broad range of patients, convenient to use, and inexpensive, with a safety profile that is complementary to existing therapies for heart failure. The most appropriate use of digoxin, however, is for patients whose symptoms persist after the administration of agents that reduce the risk of death and hospitalization (eg, ACE inhibitors and β-blockers). Digoxin is recommended for use concomitantly with diuretics, an ACE inhibitor, and a β-blocker. The finding that digoxin has little effect on survival has diminished the mandate for its early use, but it may be used early to decrease symptoms in patients who have initiated therapy with an ACE inhibitor or a β-blocker but have not yet responded symptomatically to treatment.

As with any pharmacological treatment, the tolerability of β-blockade is an abiding concern among physicians. In the short term, high-level β-blockade can depress cardiac function, causing a temporary worsening of symptoms.⁷¹ The benefits of adrenergic inhibition, on the other hand, may not become apparent for a month or more.⁷² These characteristics may be especially worrisome to the clinician who is considering β-blockade for a patient whose symptoms are mild or apparently responsive to standard therapy. In the interest of "doing no harm," the primary caregiver may be reluctant to prescribe a drug that may make the patient feel worse at first. Another practical concern is whether the extra care required to establish β-blockade is worthwhile.

While these questions are valid, clinical experience does not suggest that the risks of initiating β-blockade outweigh the long-term therapeutic benefits. As with ACE inhibitors, β-blockers may produce unwanted adverse effects that result directly from changes in neurohormonal activity. These effects occur primarily during initiation of therapy and can be associated with a period of clinical instability that may include risks of fluid retention, hypotension, and bradycardia. This period of change in clinical status can be readily managed by adjustments in background medications and usually subsides after several weeks of treatment. Key to successful introduction of β-blocker therapy and the realization of long-term benefits is the optimization of the patient's volume status and concomitant administration of heart-failure drugs before initiation of β-blockade. This early and careful monitoring for changes in patients' clinical status enabled more than 90% of patients to tolerate β-blocker initiation in the US-CHFTP.¹

Drugs that block β-receptors and α₁-receptors may produce hypotension. Thus, starting treatment with carvedilol can produce substantial vasodilation, which is usually asymptomatic but may be accompanied by dizziness or lightheadedness; this pharmacologic effect of vasodilation is generally seen within 24 to 48 hours of the first dose or increments in dose, but usually disappears on repeated dosing without any intervention.³⁸ Starting at low doses (3.125 mg twice daily) and gradually up-titrating (at 2-week intervals to a maximum of 25-50 mg twice daily) helps to manage the dizziness. Dizziness associated with β-blockers is similar to that with ACE inhibitors—the risk is greatest in patients with the lowest blood pressure but is usually self-limited. If an episode persists, the dose and timing of administration of ACE inhibitors and vasodilators or diuretics can be adjusted.

Selective and nonselective β-blockers can slow heart rate and cardiac conduction and as a result may cause bradycardia and heart block. These changes are usually asymptomatic and less likely if the dosing regimen is started low and gradually increased to the target dose.^38,73 The risk of bradycardia and heart block increases to 5% to 10% as the dose of β-blockers is progressively increased.¹ There is no uniformly accepted threshold for a minimally acceptable resting heart rate for initiation or continuation of therapy; however, multicenter trials designated at least 50 to 60 beats per minute as the minimum to consider a dosage safe to continue treatment.^1- 2,33 If the heart rate drops to fewer than 50 beats per minute or second- or third-degree heart block develops, the dose of β-blocker should be decreased, and the possibility of drug interactions should be considered; other agents that can cause bradycardia may be beneficially discontinued.^{38,67,73- 74}

Most patients—even those who experience an acute worsening of symptoms—can tolerate, and derive significant benefit from, long-term therapy.⁷⁵ In a recent observational analysis conducted in Germany, Kropff and colleagues⁷⁶ found that 94% of patients with stable heart failure could achieve therapeutic doses of carvedilol with minimum difficulty during initiation and up-titration. Only 1% of the more than 2500 patients in the study withdrew from treatment because of adverse experiences. In addition, 98% of the participating physicians rated the tolerability and feasibility of the currently accepted dosing schedule as good or excellent.

In summary, during short-term treatment, β-blockers interfere with the positive inotropic actions of endogenous catecholamines, and cardiac function may transiently decrease. Clinical improvement with β-blocker treatment is usually not seen until after 6 to 12 weeks of continuous treatment at therapeutic doses. This delay is directly related to the time required for the antagonism of the β-receptors to result in reversal of the toxic effects produced by endogenous catecholamines. Improved clinical status and the recovery of cardiac function can be sustained for very long periods. This pattern of delayed clinical improvement and reversal of hemodynamic abnormalities is similar with the use of other neurohormonal antagonists in heart failure, such as ACE inhibitors.⁴²

Compared with the proven benefits of β-blockade—slowed disease progression, improvement of symptoms, and reductions in hospitalization and death—the relative risks of establishing therapy are small. While initiating β-blockade may place additional demands on the primary care physician, the extra effort is worthwhile, especially in light of the high rates of morbidity and mortality among patients receiving standard treatment (including ACE inhibitors) alone.

In this era of managed care, the cost-effectiveness of a particular treatment has become an important issue. From a qualitative perspective, if carvedilol is able to reduce clinical progression of heart failure by 48% (P=.01) in class II patients, the cost burden should also be improved. This primary end point in the component protocol of the US-CHFTP evaluating mild heart failure was defined as (1) death due to heart failure, (2) hospitalization for heart failure, and (3) the need for sustained increase (50% increase in dose for ≥30 days) in heart failure medications. No deaths from heart failure occurred in the carvedilol group, whereas 4 placebo patients died (2 sudden death; 2 pump failure). Hospitalization for heart failure and the need to increase heart failure medications occurred less often in patients receiving carvedilol compared with patients receiving placebo and background therapy (Figure 2). In the US-CHFTP, the duration of hospital stay was decreased by 1.3 days (P=.01) for patients undergoing carvedilol treatment.⁷⁷ Furthermore, an economic model suggests that carvedilol is highly cost-effective under 2 alternative scenarios that assess the incremental cost per year of life saved (discounted). A limited-benefits scenario (benefits of carvedilol persist for 6 months only) and an extended-benefits scenario (benefits of carvedilol persist for 6 months, then decline gradually over 3 years) show that cost per life-year saved for carvedilol was $29 477 and $12 799, respectively. The cost-effectiveness of carvedilol for heart failure compares favorably with that of other generally accepted interventions for patients with heart disease.⁷⁸

This review has attempted to support the use of β-blockade in all patients with mild symptoms of systolic heart failure (unless contraindications exist) who seem to be responding to standard therapy. For compromised LV function without symptoms, no data are available on the use of β-blocker treatment; however, the results of the postinfarction trials suggest a benefit in these patients.^79- 81 The ongoing Carvedilol Post-Infarct Survival Control in Left Ventricular Dysfunction trial is designed to address the survival effects of long-term β-blockade in patients with asymptomatic LV systolic dysfunction after myocardial infarction.

It is important to remember that the sympathetic nervous system is one of the primary contributors to the onset and progression of LV systolic dysfunction, and that chronic sympathetic activation may begin before the symptoms of heart failure emerge. The mechanical and biological effects of sustained adrenergic stimulation devastate the heart in a progressive fashion, and as long as these processes remain active, the improvements seen in conventional therapy are misleading; that is, they are improvements in symptoms alone. Even the addition of ACE inhibitors to standard regimens of digoxin and diuretics does not offer optimal protection against progressive cardiac dysfunction, as evidenced by the high rate of hospitalization and death among patients receiving such treatment. Clinical trials have shown that the use of β-blockers (particularly carvedilol, bisoprolol, and long-acting metoprolol) in addition to diuretics, digitalis glycosides, and ACE inhibitors improves cardiac function, slows disease progression, and reduces morbidity and mortality in patients with systolic heart failure. In addition, treatment with carvedilol has been shown to slow disease progression even in patients with mild symptoms of systolic dysfunction. Though the initiation and establishment of β-blockade require extra care, the cost is small compared with the consequences of postponing adrenergic intervention. By adding β-blockers to the standard regimen at an early stage of treatment, caregivers may increase the chances of improving health and prolonging life in patients with systolic heart failure.

Accepted for publication September 15, 1999.

Reprints: William T. Abraham, MD, Section of Heart Failure and Cardiac Transplantation, University of Cincinnati College of Medicine, ML 0542, 231 Bethesda Ave, Cincinnati, OH 45267-0542 (e-mail: william.abraham@uc.edu).